Using ReCap Photo for Drone Site Surveys

In my job, I need to stay up on the current software, and associated workflows.  One software that has changed greatly over the last several years has been ReCap, and its companion software ReCap Photo.  In the early versions, ReCap had two workflows, one for scans and one for photogrammetry.  You might be aware that a few years ago, Autodesk split the two workflows, keeping scanning in ReCap and introduced ReCap Photo for photogrammetry.

In a previous blog post, I wrote how to take photos of an object and generate a 3D mesh.  Another application for this technology has been using drone photography to aid in site surveying.  One of my first ReCap projects was working with a customer and Autodesk to generate a Civil 3D surface from drone photographs.  

Since the introduction of ReCap Photo, I haven't really used the site survey workflow, especially connecting a model to Civil 3D.  So I wanted to see how the process has changed, some of it is the same, but there are a lot of great additions.  I learned a good bit and wanted to share that through this blog.  I did try to create my own data set, from my own drone photos.  However, I am sad to say that my drone just doesn't work well for this type of application, so I will have to use a generic dataset  I got from Autodesk.

The whole workflow starts from picking the type of project, Aerial or Object.  Creating an Object project is my normal workflow, and just allows you to add photos.  Using Aerial gives you a lot more options, which are all centered around geolocating your project.  

After picking your project type, the second step remains the same, selecting your photographs.  One way this workflow is made easier is that if the photos have a geotag in them, ReCap Photo is capable of reading that information and use it while generating the 3D model.  One oddity that I have run into is that I will get a message warning me that not all of my photos have GPS data, so my model will not be geolocated.  However, the 3D model does come out properly geolocated.

One significant change for the better is that ReCap Photo has changed the way that Ground Control Points are handled.  After your photos are selected, you can select an icon on the left of the screen to open a different window to enter the Ground Control Point data.

In the Ground Control Points window, you can designate the Coordinate System for the Ground Control Points, you can enter their values, or load them from a file.  Once the points are loaded, you can then select a row and identify the point in your photographs.

Another useful function of this interface is to toggle the display from Photos to Map.  That will then change the right side of the window to display a Bing Maps image of the site.  In my case, since my photos are geotagged, ReCap Photo will even display camera glyphs at the locations the photos were taken from.

Once all the Ground Control Point data is entered and identified in photos, you can return to the previous screen by clicking Done at the bottom of the window.  This will take you back to the photo selection screen, where you can select Create to start the project creation.

At this point, the user just has to enter a name for the project, decide on the output files, and pick a target coordinate system.

There are three possible output files, the user can select one or all three.  The RCM file is a mesh file and would be suitable for continuing to work in ReCap Photo.  The RCS output type is a point cloud and is the best type to be placed into Civil 3D or Infraworks.  The Orthophoto and elevation map options generate 2D TIFF output.  My personal habit it to generate the RCM and RCS.

For the Target Coordinate System, ReCap Photo will default to a UTM Zone.  However, I have found that those don't typically locate properly in Civil 3D or InfraWorks.  For that reason, I always change the coordinate system to the appropriate State Plane.

You will also notice that creating this model will consume a specific amount of Autodesk Cloud Credits.  ReCap Photo determines how many Cloud Credits the project will cost based on the number of photos in the project.

After defining all the selections for the output, it is just a matter of submitting the job and waiting for the 3D model to be generated.

Here is a video demonstration of the process.

I like the new additions, I think that they give you better control over the ground control points, it is easier to set the coordinate system and the output options are easier to work with.